Butane Adsorption on Silica Supported MoOx Clusters Nanofabricated

Aug 8, 2013 - I&EC Analytical Edition ..... The beam flux was determined to be F = (2.0 ± 0.1) × 1013 butane ... The initial adsorption probability,...
0 downloads 0 Views 1MB Size
Downloaded by UNIV OF WESTERN ONTARIO on August 9, 2013 | http://pubs.acs.org Publication Date (Web): August 8, 2013 | doi: 10.1021/bk-2013-1140.ch012

Chapter 12

Butane Adsorption on Silica Supported MoOx Clusters Nanofabricated by Electron Beam Lithography J. Shan,1 A. Chakradhar,1 K. Anderson,1 J. Schmidt,1 S. Dhuey,2 and U. Burghaus*,1 1Department

of Chemistry and Biochemistry, North Dakota State University, Fargo, North Dakota 58108 2Nanofabrication Facility, Molecular Foundry, LBNL, Berkeley, California 94720 *E-mail: [email protected]; Fax: 701.231.8831; URL: www.uweburghaus.us

Electron beam lithography was used to nanofabricate 77 nm Mo clusters supported on silica. Sample morphology and chemical composition were determined by scanning electron microscopy, Auger electron spectroscopy, and X-ray photoelectron spectroscopy. The chemical activity of metallic and oxidic Mo clusters towards butane adsorption was studied by thermal desorption spectroscopy and molecular beam scattering. Whereas butane adsorbs molecularly and non-dissociatively on the metallic clusters, the adsorption probabilities on MoOx are below the detection limit of uptake experiments (< 0.05).

Introduction Why Use Electron Beam Lithography in Surface Science and Catalysis? Industrial catalysts for heterogeneously catalyzed gas-surface reactions typically consist of metallic, sulfidic, or oxidic clusters supported on a porous metal oxide or graphitic support (1). In order to mimic the structure of these systems in surface science, clusters supported on single crystal supports (2–4) and © 2013 American Chemical Society In Nanotechnology for Sustainable Energy; Hu, Y., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2013.

Downloaded by UNIV OF WESTERN ONTARIO on August 9, 2013 | http://pubs.acs.org Publication Date (Web): August 8, 2013 | doi: 10.1021/bk-2013-1140.ch012

thin films (5, 6) were studied at ultra-high vacuum (UHV) conditions. The clusters for these so-called model catalysts were mostly obtained by physical vapor deposition (PVD). Although a simple and inexpensive method, the drawback of PVD is an ill-determined morphology of the system. Consequently, a significant number of work hours were spent on just characterizing the growth of clusters on supports. Therefore, today more and more nanofabrication techniques are employed in catalysis and surface science. Among other fabrication techniques, a promising approach is utilizing electron beam lithography (EBL) to nanofabricate model catalysts with an a priori knowledge about the cluster morphology. EBL allows for engineering samples with predetermined cluster size, height, rim length, dispersion, and chemical composition (7, 8). In the simplest case, EBL is used to nanofabricate a regular pattern of clusters on a support. These model systems are named model-nano-array catalysts (MNAC) (9). We report here about an example of applying EBL to study model catalysts.

Why Study Molybdenum EBL Model Catalysts? In prior studies we characterized Au, Cu, and Cu-oxide MNAC (10–17). In addition, one can find studies on Pt and Pd MNAC in the recent catalysis/surface science literature (see e.g. refs. (8, 18–23)). For this current project we looked at Mo MNAC. Mo single crystals have been studied extensively as a hydrodesulfurization (HDS) catalyst (24), Mo clusters and inorganic nanotubes were also recently considered as HDS catalysts (25–27). In regard of the topic of this meeting more interesting, however, is the observation that Mo clusters can decompose small organic molecules such as methanol (MeOH). Mo is very reactive, for example, decomposition of thiophene, CO, and CO2 was reported (28, 29). A future proposal is using MeOH as an energy carrier to generate, for example, hydrogen for fuel cell vehicles using on-board catalysts and thereby circumventing the hydrogen gas storage problem. MeOH could be synthesized in the first place from hydrogen and CO2. It has been envisioned to recycle CO2 from air and to generate hydrogen by water splitting via sunlight, forming a sustainable cycle to fuel vehicles (30). Therefore, it is interesting to characterize the surface chemistry of small organic molecules interacting with Mo-based catalysts. However, rather than going through an additional loop (first synthesizing MeOH then decomposing it), at least in the foreseeable future, it would perhaps be more efficient to decompose methane (natural gas) directly to generate hydrogen. Therefore, in this study we characterized the adsorption of alkanes on Mo clusters utilizing nanotechnology to fabricate the samples. Our motivation from an applied perspective was summarized in the preceding paragraph. Scientifically, it is interesting to compare the chemical activities of metallic and oxidic clusters. The properties of catalysts can be dramatically altered by surface oxidation (31–33). This gives one a means to tune the catalytic behavior of a given system or to prevent catalyst deactivation. To understand the underlying mechanism, molybdenum and molybdenum oxides have been extensively studied in recent years, as an important example (34–42). 296 In Nanotechnology for Sustainable Energy; Hu, Y., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2013.

Downloaded by UNIV OF WESTERN ONTARIO on August 9, 2013 | http://pubs.acs.org Publication Date (Web): August 8, 2013 | doi: 10.1021/bk-2013-1140.ch012

Brief Literature Survey – Molybdenum The decomposition of MeOH has been investigated on metallic and oxygenmodified Mo single crystals (43). Accordingly, methanol decomposes on clean Mo(100), and desorb without decomposition from oxygen-saturated Mo(100) (43, 44). The saturation coverage for MeOH on Mo(100) decreases with the extent of surface oxidation, similar to Mo(110) (see ref. (40)). These studies suggest that although Mo is quite reactive, complete oxidation makes the Mo surfaces inert. Most early studies were conducted on various Mo single crystals. More recently, however, supported Mo and Mo oxides clusters were considered (45–47). Berkó, et. al, deposited Mo clusters on TiO2(110) by PVD (46). The clusters were stable up to annealing temperatures of 900 K, and partial oxidation, probably by lattice oxygen, occurred at 900 - 1050 K. Mo clusters supported on SiO2 via PVD, were oxidized by annealing at 600 K in 1x10-6 mbar molecular oxygen (48). In contrast, oxidizing Mo single crystals requires significantly harsher reaction conditions (40, 42). Therefore, one may speculate that real-world catalysts are strongly affected by oxidation, leading to catalyst deactivation. Thus, elucidating mechanistic details also has implications for applications. Historically, Mo catalysts were used for HDS of coal (48, 49) and raw oil (24), while Mo oxides catalyze oxidations (50, 51).

Literature Survey – Alkanes Bond activation of alkanes has been studied for decades as a model system, but with a historic focus on metal single crystal surfaces (see, e.g., refs. (52–59)). Only a few surface science projects about the adsorption of alkanes on non-metallic systems have been conducted. For MgO (60), ZnO (61), rutile TiO2 (62, 63), silica (64, 65), and graphitic systems (66–68), only molecular adsorption is seen. Recently, for an alkaline earth metal oxide single crystal, CaO(100) (ref. (69)), bond activation of butane was evident. Later, bond cleavage on transition metal oxides, PdO thin films, was reported (70). At present, the only metal oxide single crystals/thin films studied in more detail with surface science techniques that promote alkane dissociation are Pd and Ca oxides. Palladium is quite rare and expensive and therefore not the best choice for a sustainable economy. Note that bond activation of alkanes was also present for anatase TiO2 thin films (, probably the first system studied in this regard (63)), however, here the alkanes decompose entirely, which makes detailed characterization very cumbersome. In addition, the TiO2 catalyst is poisoned in that process. Silica is often used as a support for nanostructures, as also in our project, since it is cheap and quite inert. For example, CO basically does not adsorb on silica (11, 71). Physisorption of alkanes on silica, however, was seen before (64, 65). The earlier project was motivated by providing blind experiments for surface science studies on carbon nanotubes (CNTs); here CNTs are typically drop-and-dried on supports (at ambient pressure). Therefore, in the prior study, the silica samples were not UHV cleaned and probably covered with a carbon layer. Graphitic systems adsorb alkanes (67, 72). 297 In Nanotechnology for Sustainable Energy; Hu, Y., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2013.

In this report, the change in catalytic activity while oxidizing Mo clusters is characterized using an alkane (butane) as the probe molecule. The model catalyst was nanofabricated by electron beam lithography. As a main conclusion: metallic Mo clusters show some catalytic activity, whereas Mo oxides are completely inert towards alkane adsorption. Perhaps amazingly, even the metallic phase does not lead to bond activation in butane, although Mo decomposes a variety of other small organic molecules (43, 44, 73).

Downloaded by UNIV OF WESTERN ONTARIO on August 9, 2013 | http://pubs.acs.org Publication Date (Web): August 8, 2013 | doi: 10.1021/bk-2013-1140.ch012

Experimental Setup The measurements were conducted by a home-built, triply differentially pumped molecular beam scattering system (14). The supersonic molecular beam is attached to a scattering chamber, which houses two quadrupole mass spectrometers (QMS). One spectrometer is mounted perpendicular to the beam direction (for beam scattering measurements), and the other one is aligned with the direction of the beam (for time-of-flight (TOF) measurements). A double pass cylindrical mirror analyzer equipped with an electron gun is used for Auger electron spectroscopy (AES). In combination with an X-ray source, this analyzer is also utilized for X-ray photoelectron spectroscopy (XPS). In addition, a home-built metal evaporator, a commercial electron beam metal evaporator, a sputter gun, an atomic hydrogen source, and several leak valves are mounted on the scattering chamber. In molecular beam and thermal desorption spectroscopy (TDS) experiments, n-butane was dosed with the beam on the sample, in order to suppress sample holder effects. The impact energy of n-butane was varied within Ei = 0.1 - 1.2 eV by using a pure beam or by seeding with Helium, combined with a variation of the nozzle temperature in the range of 300 - 750 K. The impact energies of n-butane were measured by TOF. The beam flux was determined to be F = (2.0 ± 0.1) × 1013 butane molecules cm-2 s-1 for the pure butane beam. The initial adsorption probability, S0, of n-butane on Mo clusters was obtained by King and Wells type uptake experiments (74) directing the beam perpendicular to the surface plane. For a single measurement, the uncertainty in S0, as estimated from the signal-to-noise ratio, amounts to ± 0.05. The sample could be heated to 1000 K by electron bombardment and cooled down to 85 K by bubbling He gas through a liquid nitrogen containing dewar. The temperature was read by a K-type thermocouple. The reading of the thermocouple was calibrated in situ within ± 5 K by recording the condensation peaks of nbutane in TDS experiments. The heating rate amounted to 1.7 K/s. TDS data were collected with a shielded QMS. In TDS measurements, the distance from the sample to the shield amounts to only ~1 mm. For AES, the electron energy amounts to 2 keV with a modulation voltage of 2 eV. For XPS, the Mg Kα line (at 1253.6 eV) was used with pass energy of 50 eV of the electron analyzer. The XPS spectra were referenced with respect to the O 1s line at a binding energy of 532.9 eV (see ref. (75)). Uncertainties reported for AES intensities are based on the signal-to-noise ratio of the spectra. 298 In Nanotechnology for Sustainable Energy; Hu, Y., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2013.

Scanning electron microscopy (SEM) images of the samples were collected before the UHV experiments at Lawrence Berkeley National Laboratory with a Zeiss Ultra 60 equipped with a field emission gun. In addition, SEM imaging after the experiments was conducted at Brookhaven National Laboratory with a Hitachi S-4800 UHR.

Results and Discussion

Downloaded by UNIV OF WESTERN ONTARIO on August 9, 2013 | http://pubs.acs.org Publication Date (Web): August 8, 2013 | doi: 10.1021/bk-2013-1140.ch012

Sample Morphology The EBL sample was nanofabricated at Molecular Foundry (Lawrence Berkeley National Laboratory) (12). Mo clusters with a diameter of ds = (77 ± 2) nm were arranged in a hexagonal pattern with a lattice constant of as = 150 nm, and a height of about 5 nm. A 5 mm by 5 mm area was covered with the Mo clusters on a 10 by 10 mm silica support. Although the Mo cluster size, shape, and height, as well as lattice constant were predetermined in the EBL fabrication, the sample was still inspected by scanning electron microscopy (SEM), see Fig. 1A. The size distribution of the clusters is depicted in Fig. 1C as a bar diagram and was determined by using a commercial software tool (Pixcavator). The cluster size distribution is with 2/77 = 2.5% narrow (, using the FWHM of the Gaussian fit shown in Fig. 1C as a solid line). The Mo coverage (total Mo area vs. support area) is calculated as 0.25 ML. (A monolayer, 1 ML, corresponds to a completely covered surface.) Fig. 1B depicts SEM images collected after the UHV experiments. Evidently the sample’s morphology was conserved (cf., Fig. 1A and 1B). The cluster size distribution (shown in Fig. 1D) was also unaltered, i.e., significant sintering was absent. In order to better compare the cluster size distributions, Fig. 1C depicts the Gaussian fits together. A slight broadening of the distribution was evident for the used model catalyst. This may indicate the capture of small metal residuals from the lift-off process (or along the clusters rim) by the larger Mo clusters, cause by the sample annealing. Given the size of the clusters, a relaxation of grain boundaries appears unlikely. Sample Cleaning – AES and XPS Characterization Cleaning EBL samples without destroying the morphology or irreversibly changing the chemical composition is non-trivial (76, 77) and therefore described in more detail. Mo single crystals were usually cleaned by a thermal treatment at rather high temperature, e.g., annealing at ~1700 K under O2 atmosphere followed by flashing to 2300 K was used (34–42). Unfortunately, this standard cleaning procedure is not applicable to EBL samples. Such a high temperature would likely result in cluster sintering, since the Tammann temperature of Mo may amount to only ~1450 K, assuming that it equals half of the bulk melting temperature. Note that, sintering of Mo clusters on TiO2 and SiO2 supports, respectively, has been observed experimentally already upon annealing at ~ 1050 K (refs. (46, 78)). Therefore, the Mo EBL sample studied here was initially cleaned by mild room temperature sputtering/annealing cycles (one minute each, sample current 2 μA), 299 In Nanotechnology for Sustainable Energy; Hu, Y., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2013.

Downloaded by UNIV OF WESTERN ONTARIO on August 9, 2013 | http://pubs.acs.org Publication Date (Web): August 8, 2013 | doi: 10.1021/bk-2013-1140.ch012

followed by annealing at 700 K in oxygen (1 × 10-8 mbar), and a flash to only 750 K. During this cleaning procedure, the sample was inspected regularly by AES (Fig. 2A). The sputter cycles were continued up to a C/Mo AES peak ratio of 0.2. Thus, the C impurities did act as a buffer layer protecting the Mo clusters. A similar procedure was used before while cleaning thin film samples transferred through air (63). Afterwards, we switched from a “physical cleaning” to a “chemical treatment”: cycles of oxygen annealing at 700 K for 10 minutes were applied. Finally carbon containing impurities were below the detection limit of the AES system (Fig. 2B). Unfortunately, this procedure will also likely oxidize the Mo clusters (48). Again, the high-temperature (1050 K) (79) annealing used to reduce Mo single crystals does not work for EBL clusters. Therefore, as a final cleaning step, the EBL sample was annealed in a flux of atomic hydrogen (p(H2) = 1 × 10-7 mbar, Ts = 750 K, for two hours). According to prior work in our group, this will successfully reduce the oxides (48, 80). Note that during the whole cleaning procedure, the Mo/Si AES intensity ratio remained constant and sample morphology was conserved (Fig. 1).

Figure 1. SEM images of the Mo EBL clusters supported on silica (A) before and (B) after the UHV experiments. The corresponding cluster size distributions are given in panel (C) and (D). Depicted is the cluster number (normalized with respect of the total cluster number) vs. the clusters’ diameter. The lines are Gaussian fits of the cluster size distributions before (red dotted line) and after (blue solid line) the UHV experiments. 300 In Nanotechnology for Sustainable Energy; Hu, Y., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2013.

Downloaded by UNIV OF WESTERN ONTARIO on August 9, 2013 | http://pubs.acs.org Publication Date (Web): August 8, 2013 | doi: 10.1021/bk-2013-1140.ch012

Figure 2. Cleaning procedure of the Mo EBL sample. (A) cleaning with Ar+ sputtering, (B) cleaning with O2 annealing. Fig. 3A depicts examples of AES survey scans of the as received and cleaned Mo EBL sample, respectively. Besides Si and O AES peaks from the support as well as Mo AES features (81), no other structures were evident for the cleaned sample. Similarly, a XPS survey scan of the cleaned Mo EBL sample is given in Fig. 3B. The spectra confirm that no carbon or other impurities are present on the cleaned surface. The XPS peak positions of Si, O, and Mo agree with reference data (75). The inset depicts the silica XPS 2p region. For fully oxidized silica only one peak is present, whereas partially reduced silica shows two peaks (75). Thus, 301 In Nanotechnology for Sustainable Energy; Hu, Y., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2013.

Downloaded by UNIV OF WESTERN ONTARIO on August 9, 2013 | http://pubs.acs.org Publication Date (Web): August 8, 2013 | doi: 10.1021/bk-2013-1140.ch012

the cleaning procedure did not reduce the silica support (80, 82). This fact may be important, since silica is non-reactive, but silicon is highly reactive (64, 65).

Figure 3. (A) AES spectra of the Mo EBL sample before cleaning (dashed line) and after cleaning (solid line). (B) XPS spectra of the Mo EBL sample after cleaning. The inset shows the XPS Si 2p region.

302 In Nanotechnology for Sustainable Energy; Hu, Y., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2013.

Both metallic and oxidic Mo clusters were considered in this study. Unfortunately, metallic and oxidic molybdenum have very similar XPS signatures (42, 83, 84). However, according to refs. (34, 40–42, 48, 79, 83) oxygen and hydrogen annealing, respectively, can be used to change the oxidation state of molybdenum. Therefore, Mo oxides were reduced in a flux of atomic hydrogen (Ts = 750 K, t = 2 h), while metallic clusters were oxidized by annealing in an oxygen ambient (p = 1 × 10-6 mbar, 750 K, 20 min).

Downloaded by UNIV OF WESTERN ONTARIO on August 9, 2013 | http://pubs.acs.org Publication Date (Web): August 8, 2013 | doi: 10.1021/bk-2013-1140.ch012

Adsorption Dynamics - Typical Adsorption Transients The adsorption dynamics of n-butane on the supported Mo clusters were characterized by molecular beam scattering. The solid lines in Fig. 4 show typical adsorption transient on Mo clusters, while the dashed lines depict results for oxidic clusters. Results of several experiments at identical measuring conditions are shown in order to illustrate the reproducibility of the measurements. Displayed is the exposure time (t), vs. the partial pressure of n-butane in the scattering chamber. At t = 0 s, a beam flag is opened and n-butane molecules start to strike the surface. If n-butane molecules do not adsorb on the surface, the transient resembles only a step function due to the increase in the equilibrium pressure. However, if the probe molecules adsorb, the alkane’s partial pressure initially increases fast, but approaches the saturation level slowly hereafter. Once the surface is saturated, all butane molecules will be backscattered since thermal desorption and butane condensation can be neglected at the chosen adsorption temperature (Ts = 115 K). Saturation was reached for t > 10 s. The area above the transient and below the saturation level equals the number of adsorbed molecules. Therefore, it is evident from Fig. 4 that the alkane does not build up any significant concentration on the silica supported oxidic clusters, but does adsorb on supported metallic Mo clusters. From the known geometrical size of the Mo clusters and the lattice constant of the array catalyst (see Fig. 1), the Mo coverage can be calculated as 0.25 ML. Since the surface density of a typical Mo surface amounts to 1.4e15 atom cm-2 (ref. (85)), the atomic density of the Mo clusters supported on SiO2 reads 3.5e14 atom cm-2. By integrating the transients shown in Fig. 4, we conclude that the total butane molecular density on the Mo clusters at saturation amounts to 1.0e14 molecule cm-2. In other words, at saturation coverage the ratio of butane molecules to Mo atoms is only about 0.3, i.e., approximately 3.5 Mo atom can accommodate one butane molecule. That result appears reasonable considering the size of n-butane (5.5 Å). Furthermore, the adsorption probability, S(t), of n-butane can be obtained from these adsorption transients (86). In Fig. 4, the curves are normalized such that 1- S(t) vs. t is depicted. Thus, the adsorption probability can directly be read from these graphs. For example, the initial adsorption probability, S0, of pure n-butane can be determined at t = 0.

303 In Nanotechnology for Sustainable Energy; Hu, Y., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2013.

Downloaded by UNIV OF WESTERN ONTARIO on August 9, 2013 | http://pubs.acs.org Publication Date (Web): August 8, 2013 | doi: 10.1021/bk-2013-1140.ch012

Figure 4. Typical adsorption transients of n-butane on metallic Mo clusters (solid lines), and on oxidic Mo clusters (dotted lines). Energy Dependence of the Initial Adsorption Probability of n-Butane In Fig. 5, S0 on metallic Mo clusters is shown as a function of impact energy, Ei. The sample temperature was kept constant at 115 K. It is evident that S0 decreases with increasing Ei. This trend reflects the decreasing efficiency of gas-tosurface energy transfer with increasing Ei. In simple terms, the larger Ei, the larger the speed of the molecules when impinging onto the surface and the smaller the interaction time with the surface. Therefore, the efficiency of the energy transfer processes of n-butane molecules decreases with Ei, hence S0 decreases with Ei. This observation is consistent with non-activated and molecular adsorption of nbutane on metallic Mo clusters as is the absence of carbon after several adsorption/ desorption cycles, evident from the AES/XPS experiments. Kinetics of n-Butane on Mo Clusters Fig. 6 summarizes butane TDS data on metallic Mo clusters (solid lines), as well as on oxidic Mo clusters (dashed line). Pure butane was dosed with the molecular beam system at Ts = 90 K. Therefore, the exposure,χ, is given as an exposure time in sec. We used exposures up to 30 sec for TDS. With the measured beam flux of F = (2.0 ± 0.1) × 1013 butane molecules cm-2 s-1 that translates into a maximum exposure of ~ 1 Langmuir. For the oxidized Mo clusters, no desorption of butane is seen. These data also suggest that the adsorption of n-butane from the silica support is below the detection level. This is consistent with the adsorption transient of butane on the oxidized sample (Fig. 4). 304 In Nanotechnology for Sustainable Energy; Hu, Y., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2013.

Downloaded by UNIV OF WESTERN ONTARIO on August 9, 2013 | http://pubs.acs.org Publication Date (Web): August 8, 2013 | doi: 10.1021/bk-2013-1140.ch012

Figure 5. Initial adsorption probability of n-butane on metallic Mo clusters as a function of impact energy.

Figure 6. TDS of n-butane on metallic Mo clusters (solid lines), and on oxidic Mo clusters (dotted lines). For metallic clusters, at the smallest exposure of χ = 2 s, two TDS features are evident, one centered at 225 K (α peak), and the other centered at 160 K (β peak). With increasing χ, both features shift to lower temperatures. At the largest 305 In Nanotechnology for Sustainable Energy; Hu, Y., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2013.

Downloaded by UNIV OF WESTERN ONTARIO on August 9, 2013 | http://pubs.acs.org Publication Date (Web): August 8, 2013 | doi: 10.1021/bk-2013-1140.ch012

exposure χ = 30 s, the α peak shifts to 190 K while the β peak shift to 140 K. For χ = 10 s, a new feature appears, first as a shoulder at 105 K. This shoulder develops into a distinct peak at χ = 30 s. The low temperature onsets of the γ peaks line up, which is characteristic of condensed alkanes. Note also that, according to the beam scattering experiments (Fig. 4), the surface saturates for exposures of about 10 sec. Indeed, at about that exposure the γ peak develops in TDS. The α and β peaks are related to the desorption of butane from the metallic Mo clusters; we assign these features to different types of adsorption sites. For example, the higher temperature structure is related to desorption from defect sites (such as the rim of the clusters), while the low temperature feature to pristine sites (terrace sites) (80, 87). The shift of α and β peak positions with butane exposure are likely due to repulsive lateral interactions among butane molecules. Due to the overlapping peaks, the magnitude of the peak shift is difficult to quantify. Repulsive interactions among alkanes have been seen before and were described by polarization effects induced by the support (88). Fig.7 shows multi-mass TDS data of n-butane on metallic Mo clusters (red bar), as well as the mass spectra of gaseous butane (gray bar), both detected with the same mass spectrometer. In the multi-mass TDS experiments, a constant exposure of χ = 5s was used. The peak intensities in the TDS spectra are normalized with respect to the intensity at m/e = 43 (where the strongest signal of n-butane in detected). Fig. 7 clearly shows that the mass scans for gaseous butane match those of the multi-mass TDS data. This indicates that only a molecular adsorption pathway is present. That conclusion is consistent with the absence of carbon in AES/XPS after TDS experiments, as well as the energy dependence of the initial adsorption probability.

Figure 7. Multi-mass TDS of n-butane on metallic Mo clusters comparing with a mass scan of gaseous n-butane. 306 In Nanotechnology for Sustainable Energy; Hu, Y., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2013.

Downloaded by UNIV OF WESTERN ONTARIO on August 9, 2013 | http://pubs.acs.org Publication Date (Web): August 8, 2013 | doi: 10.1021/bk-2013-1140.ch012

Conclusions According to the molecular beam scattering data, n-butane adsorbs on silica supported Mo clusters with initial adsorption probabilities, S0, within the range of S0 = 0.60 – 0.32, depending on impact energy, whereas on oxidic clusters S0 was below the detection limit of S0 = 0.05. Thus, the metallic clusters are reactive, but the oxidic clusters are not. Based on XPS and AES data, the adsorption is molecular, i.e., in both cases bond activation of the alkane was absent. TDS results consistently confirm adsorption on the silica supported Mo clusters. In contrast, even large exposures on the oxidic cluster sample did not allow for detecting nbutane desorption in TDS. Note that the silica support was UHV cleaned and the alkane was dosed onto the surface using a collimated molecular beam scattering system. Similar results have been reported before: metallic Mo surfaces decompose efficiently methanol (34, 40) and thiophene (48), whereas oxidic Mo surfaces are rather inert. The crystal structure was invoked as a plausible explanation (89). For example, MoO3 has an orthorhombic layered structure, consisting of MoO6 octahedra, forming the basal plane of MoO3 single crystals. That surface is oxygen terminated and therefore expected to be rather inert. The oxide formation would be one mechanism for catalyst deactivation. However, it perhaps remains surprising that even the metallic Mo clusters are catalytically not very active, since only molecular adsorption of the alkane was evident. The advantage of using EBL samples for this project, in particular, is the a priori knowledge for samples’ morphology. Thus, a time consuming morphology characterization is unnessesary. A detailed discussion of this approach in general can be found in ref. (9). In ref. (14, 15), the comparison of different cluster sizes, for example, allows to identify the clusters’ rim as the active site of surface processes using simple kinetics techniques.

Acknowledgments Discussions with Stephano Cabrini at Molecular Foundry are acknowledged and SEM images were collected by Ming Lu at Brookhaven national labs. Financial support was provided by a NSF-CAREER award (CHE-0743932).

References 1. 2. 3. 4. 5. 6. 7.

Thomas, J. M.; Thomas, W. J. Principles and Practice of Heterogeneous Catalysis; VCH: 2005. Choudhary, T. V.; Goodman, D. W. Top. Catal. 2002, 21, 25. Goodman, D. W. J. Chem. Phys. 1996, 100, 13090. Henry, C. R. Surf. Sci. Rep. 1998, 31, 235. Freund, H. J. Angew. Chem. 1997, 109, 444. Freund, H. J. Surf. Sci. 2007, 601, 1438–1442. Avoyan, A.; Rupprechter, G.; Eppler, A. S.; Somorjai, G. A. Top. Catal. 2000, 10, 107–113. 307 In Nanotechnology for Sustainable Energy; Hu, Y., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2013.

8. 9. 10. 11.

Downloaded by UNIV OF WESTERN ONTARIO on August 9, 2013 | http://pubs.acs.org Publication Date (Web): August 8, 2013 | doi: 10.1021/bk-2013-1140.ch012

12. 13. 14. 15. 16. 17.

18. 19. 20. 21. 22. 23. 24. 25.

26.

27. 28. 29.

30. 31.

Johanek, V.; Laurin, M.; Grant, A. W.; Kasemo, B.; Henry, C. R.; Libuda, J. Science 2004, 304, 1639. Burghaus, U. In Catalysis Series; Spivey, J. J., Ed.; Royal Society of Chemistry: Cambridge, UK, in press. Kadossov, E.; Justin, J.; Lu, M.; Rosenmann, D.; Ocola, L. E.; Cabrini, S.; Burghaus, U. Chem. Phys. Lett. 2009, 483, 250–253. Kadossov, E.; Cabrini, S.; Burghaus, U. J. Mol. Catal. A: Chem. 2010, 321, 101–109. Komarneni, M.; Shan, J.; Chakradhar, A.; Kadossov, E.; Cabrini, S.; Burghaus, U. J. Phys. Chem. C 2011, 116, 5792–5801. Shan, J.; Komarneni, M.; Burghaus, U. Chem. Phys. Lett. 2011, 517, 59–61. Chakradhar, A.; Shan, J.; Komarneni, M. R.; Lu, M.; Burghaus, U. Chem. Phys. Lett 2012, 544, 70–72. Shan, J.; Chakradhar, A.; Komarneni, M.; Burghaus, U. J. Chem. Phys. C 2012, 116, 18930–18936. Kadossov, E.; Burghaus, U. Catal. Lett. 2010, 134, 228–232. Shan, J.; Chakradhar, A.; Komarneni, M.; Kadossov, E.; Schmidt, J.; Burghaus, U.; Rosenmann, D.; Ocola, L. E.; Dhuey, S.; Cabrini, S. Am. Chem. Soc., Div. Fuel Chem. 2012, in press. Grunes, J.; Zhu, J.; Anderson, E. A.; Somorjai, G. A. J. Phys. Chem. B 2002, 106, 11463. Grunes, J.; Zhu, J.; Yang, M.; Somorjai, G. A. Catal. Lett. 2003, 4, 157. Jacobs, P. W.; Wind, S. J.; Ribeiro, F. H.; Somorjai, G. A. Surf. Sci. 1997, 372, L249. Johanek, V.; Laurin, M.; Hoffmann, J.; Schauermann, S.; Grant, A. W.; Kasemo, B.; Libuda, J.; Freund, H. J. Surf. Sci. 2004, 561, L218–L224. Laurin, M.; Johanek, V.; Grant, A. W.; Kasemo, B.; Libuda, J.; Freund, H. J. J. Chem. Phys. 2005, 122, 084713. Laurin, M.; Johanek, V.; Grant, A. W.; Kasemo, B.; Libuda, J.; Freund, H. J. J. Chem. Phys. 2005, 123, 054701. Topsoe, H.; Clausen, B. S.; Massoth, F. E. J. Catal. 1996, 163, 409. Lauritzen, J. V.; Nyberg, M.; Vang, R. T.; Bollinger, M. V.; Clausen, B. S.; Topsoe, H.; Jacobsen, K. W.; Laegsgaard, E.; Norskov, J. K.; Besenbacher, F. Nanotechnology 2003, 14, 385–389. Burghaus, U. Gas-carbon nanotubes interactions: a review of ultra-high vacuum surface science studies on CNTs. Carbon nanotubes - Research Trends; Nova Science: New York, 2009. Burghaus, U.; Zak, A.; Rosentsveig, R. Isr. J. Chem. 2010, 50, 449–452. Lopez-Carreno, L. D.; Heras, J. M.; Viscido, L. Surf. Sci. 1997, 377-379, 615. Jaworowski, A. J.; Smedh, M.; Borg, M.; Sandell, A.; Beutler, A.; Sorensen, S. L.; Lundgren, E.; Andersen, J. N. Surf. Sci. 2001, 492, 185–194. Olah, G. A.; Goeppert, A.; Prakash, G. K. S., 2003, Beyond oil and gas: the methanol economy, Wiley-VCH, 2003, ISBN 0-471-41782-3. Evans, U. R. The corrosion and oxidation of metals: scientific principles and practical applications; St. Martin’s Press: 1960. 308 In Nanotechnology for Sustainable Energy; Hu, Y., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2013.

Downloaded by UNIV OF WESTERN ONTARIO on August 9, 2013 | http://pubs.acs.org Publication Date (Web): August 8, 2013 | doi: 10.1021/bk-2013-1140.ch012

32. Bond, G. C. Heterogeneous catalysis: principles and applications; Oxford University Press: 1974. 33. Kadossov, E.; Goering, J.; Burghaus, U. Surf. Sci. 2008, 602, 811–818. 34. Miles, S. L.; Bernasek, S. L.; Gland, J. L. J. Phys. Chem. 1983, 87 (9), 1626–1630. 35. Sierka, M.; Todorova, T. K.; Sauer, J.; Kaya, S.; Stacchiola, D.; Weissenrieder, J.; Shaikhutdinov, S.; Freund, H. J. J. Chem. Phys. 2007, 126 (23), 234710 . 36. Ko, E. I.; Madix, R. J. Surf. Sci. 1981, 112 (3), 373–385. 37. Deiner, L. J.; Kang, D. H.; Friend, C. A. J. Phys. Chem. B 2005, 109 (26), 12826–12831. 38. Deiner, L. J.; Kang, D. H.; Friend, C. M. J. Phys. Chem. B 2005, 109 (16), 8085–8090. 39. Deiner, L. J.; Chan, A. S. Y.; Sheehy, M. A.; Friend, C. M. Surf. Sci. 2004, 555 (1−3), L127–L132. 40. Street, S. C.; Liu, G.; Goodman, D. W. Surf. Sci. 1997, 385 (2−3), L971–L977. 41. Sasaki, T.; Goto, Y.; Tero, R.; Fukui, K.; Iwasawa, Y. Surf. Sci. 2002, 502, 136–143. 42. Kaya, S.; Weissenrieder, J.; Stacchiola, D.; Todorova, T. K.; Sierka, M.; Sauer, J.; Shaikhutdinov, S.; Freund, H. J. Surf. Sci. 2008, 602 (21), 3338–3342. 43. Miles, S. L.; Bernasek, S. L.; Gland, J. L. J. Phys. Chem. 1983, 87, 1626–30. 44. Ko, E. I.; Madix, R. J. Surf. Sci. 1981, 109, 221. 45. Chun, W. J.; Asakura, K.; Iwasawa, Y. Chem. Phys. Lett. 1998, 288 (5−6), 868–872. 46. Berko, A.; Magony, A.; Szoko, J. Langmuir 2005, 21 (10), 4562–4570. 47. Kitchin, J. R.; Barteau, M. A.; Chen, J. G. G. Surf. Sci. 2003, 526 (3), 323–331. 48. Komarneni, M.; Kadossov, E.; Justin, J.; Lu, M.; Burghaus, U. Surf. Sci. 2010, 604 (13−14), 1221–1229. 49. Kibsgaard, J.; Lauritsen, J. V.; Laegsgaard, E.; Clausen, B. S.; Topsoe, H.; Besenbacher, F. J. Am. Chem. Soc. 2006, 128 (42), 13950–13958. 50. Ozkan, U.; Schrader, G. L. J. Catal. 1985, 95 (1), 120–136. 51. Tong, Y. D.; Lunsford, J. H. J. Am. Chem. Soc. 1991, 113 (13), 4741–4746. 52. Mullins, C. B.; Weinberg, W. H. Springer Series in Surface Science; Springer GmbH: 1994; Vol. 34, Ch. 7. 53. Weinberg, W. H.; Sun, Y. K. Science 1991, 253, 542. 54. Weaver, J. F.; Carlsson, A. F.; Madix, R. J. Surf. Sci. Rep. 2003, 50, 107–199. 55. Zaera, F. Catal. Today 2003, 81, 149–157. 56. Juurlink, L. B. F.; Killelea, D. R.; Utz, A. L. Prog. Surf. Sci. 2009, 84, 69. 57. Wodtke, A. M.; Matsiev, D.; Auerbach, D. J. Prog. Surf. Sci. 2008, 83. 58. Ding, J.; Burghaus, U.; Weinberg, W. H. Surf. Sci. 2000, 446, 46. 59. Larsen, J. H.; Chorkendorff, I. Surf. Sci. Rep. 1999, 35, 163–222. 60. Tait, S. L.; Dohnalek, Z.; Campbell, C. T.; Kay, B. D. J. Chem. Phys. 2005, 122, 164708. 61. Wang, J.; Hokkanen, B.; Burghaus, U. Surf. Sci. 2005, 577, 158–166. 309 In Nanotechnology for Sustainable Energy; Hu, Y., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2013.

Downloaded by UNIV OF WESTERN ONTARIO on August 9, 2013 | http://pubs.acs.org Publication Date (Web): August 8, 2013 | doi: 10.1021/bk-2013-1140.ch012

62. Hokkanen, B.; Funk, S.; Burghaus, U.; Ghicov, A.; Schmuki, P. Surf. Sci. 2007, 601, 4620–4628. 63. Goering, J.; Kadossov, E.; Burghaus, U.; Yu, Z. Q.; Thevuthasan, S.; Saraf, L. V. Catal. Lett. 2007, 116, 9. 64. Funk, S.; Nurkic, T.; Burghaus, U. Appl. Surf. Sci. 2007, 253, 4860–4865. 65. Funk, S.; Goering, J.; Burghaus, U. Appl. Surf. Sci. 2008, 254, 5271–5275. 66. Paserba, K. R.; Gellman, A. J. Phys. Rev. Lett. 2001, 86, 4338. 67. Kadossov, E.; Goering, J.; Burghaus, U. Surf. Sci. 2007, 601, 3421–3425. 68. Funk, S.; Hokkanen, B.; Nurkig, T.; Burghaus, U.; White, B.; O’Brien, S.; Turro, N. J. Phys. Chem. C 2007, 111, 8043–8049. 69. Chakradhar, A.; Liu, Y.; Schmidt, J.; Kadossov, E.; Burghaus, U. Surf. Sci. 2011, 605, 1534–1540. 70. Weaver, J. F.; Devarajan, S. P.; Hakanoglu, C. J. Phys. Chem. C 2009, 113, 9773–9782. 71. Gross, E.; Asscher, M.; Lundwall, M.; Goodman, D. W. J. Phys. Chem. C 2007, 111, 16197–16201. 72. Gellman, A. J.; Paserba, K. P. J. Phys. Chem. B 2002, 106 (51), 13231. 73. Komarneni, M.; Kadossov, E.; Justin, J.; Ming, M.; Burghaus, U. Surf. Sci. 2010, 604, 1221–1229. 74. King, D. A.; Wells, M. G. Surf. Sci. 1972, 29, 454. 75. Moulder, J. F.; Stickle, W. F.; Sobol, P. E.; Bomben, K. D. Handbook of x-ray photoelectron spectroscopy; Perkin-Elmer-Corp: USA, 1980. 76. Ribeiro, F. H.; Somorjai, G. A. Recl. Trav. Chim. Pays-Bas 1994, 113 (10), 419–422. 77. Wong, K.; Johansson, S.; Kasemo, B. Faraday Discuss. 1996, 105, 237–246. 78. Wang, F. P.; Cheng, L. S.; Wang, P. X.; Lu, K. Q. Thin Solid Films 2000, 379 (1-2), 308–312. 79. Kim, B. S.; K., E. Y.; Jeon, H. S.; Lee, H. I.; Lee, J. C. Mater. Trans. 2008, 49 (9), 2147–2152. 80. Komarneni, M.; Shan, J.; Chakradhar, A.; Kadossov, E.; Cabrini, S.; Burghaus, U. J. Phys. Chem. C 2012, 116 (9), 5792–5801. 81. Hedberg, C. L. Handbook of Auger Electron Spectroscopy; Physical Electronics, Inc: USA, 1970. 82. Komarneni, M.; Shan, J.; Burghaus, U. J. Phys. Chem. C 2011, 115 (33), 16590–16597. 83. Schroeder, T.; Zegenhagen, J.; Magg, N.; Immaraporn, B.; Freund, H. J. Surf. Sci. 2004, 552 (1−3), 85–97. 84. Marin-Flores, O.; Scudiero, L.; Ha, S. Surf. Sci. 2009, 603 (15), 2327–2332. 85. Zhang, X.; Linsebigler, A.; Heiz, U.; Yates, J. T. J. Phys. Chem. 1993, 97, 5074–5079. 86. King, D. A.; Wells, M. G. Surf. Sci. 1972, 29 (2), 454. 87. Chakradhar, A.; S., J.; Komarneni, M.; Ming, L.; Burghaus, U. Chem. Phys. Lett. 2012, in press. 88. Kadossov, E.; Burghaus, U. Surf. Sci. 2009, 603, 2494–2501. 89. Henrich, V. E.; Cox, P. A. The Surface Science of Metal Oxides; Cambridge University Press: New York, USA, 1994. 310 In Nanotechnology for Sustainable Energy; Hu, Y., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2013.